The Coming Global Climate–Technology Revolution

Journal of Economic Perspectives—Volume 23, Number 2—Spring 2009—Pages 53–75

Scott Barrett

uppose that averting dangerous climate change meant limiting the concentration of greenhouse gases (measured in parts per million, or S ppm, by volume) in the atmosphere so that the world faced a temperature increase of no more than 2°C—a goal endorsed by the European Union. What would it take to meet this goal? Because of “climate uncertainty,” we cannot be sure. We can meet the goal with probability greater than 90 percent by limiting concentrations to 350 ppm carbon dioxide “equivalent” (the concentration of carbon dioxide that would cause the same amount of “radiative forcing” as a given mixture of carbon dioxide, or CO2, and other greenhouse gases), but we have already overshot that level (Anderson and Bows, 2008, p. 2). We can meet the goal with probability close to 50 percent by stabilizing concentrations at 450 ppm CO2 equivalent, but to do that will require that global net emissions (additions to the atmosphere minus subtractions) peak by around 2015, decline rapidly after that time, and reach zero soon after 2050. We can abandon the 2°C target and accept the likelihood of greater climate change; but stabilization at some other level, like 550, or 650, or even 750 ppm of CO2 equivalent will also require radical reductions in emissions.

Emissions of CO2 and other greenhouse gases can be reduced signiﬁcantly using existing technologies, but stabilizing concentrations will require a technological revolution—a “revolution” because it will require fundamental change, achieved within a relatively short period of time. y Scott Barrett is Lenfest–Earth Institute Professor of Natural Resource Economics, School of International and Public Affairs, Columbia University, New York, New York. At the time this article was written, he was Professor of Environmental Economics and International Political Economy, Paul H. Nitze School of Advanced International Studies, Johns Hopkins University, Washington, D.C. 54 Journal of Economic Perspectives

Inspiration for a climate–technology revolution is often drawn from the Apollo space program or the Manhattan Project, but averting dangerous climate change cannot be “solved” by a single new technology, deployed by a single government. The technological changes needed to address climate change fundamentally will have to be pervasive; they will have to involve markets; and they will have to be global in scope. Bringing about such changes involves several interlocking elements. Most importantly, a price must be put on the externality of (net) greenhouse gas emissions—in short, a carbon price. This is essential not only to create incentives for reductions in emissions but also for the private sector to innovate. Financing is also required for R&D of the fundamental kind that is not re- warded by the patent system (Arrow et al., 2008). As new technologies are developed, practical issues will arise in their dissemination. Some of these will involve intellectual property or technical expertise, both within economies and across national borders. Others will involve right-of-way and not-in-my-backyard disputes related to a new infrastructure, such as would arise for the transmission of pressurized, liquid CO2. Issues of network markets may also arise, where the market struggles to settle on a certain technology standard. Finally, some of the new technologies will introduce new risks of their own (like storage of CO2 underground) and these risks will also have to be managed. To understand all these inter-linkages and to steer the evolution of the climate–technology revolution, we need to take technology out of the black box. Individual countries can undertake all of these steps, but climate change is a global problem. Pricing carbon and developing and disseminating new technologies will require multilateral cooperation between governments working together with the private sector. It is the need for international cooperation that makes the climate–technology revolution unprecedented. My focus in this paper is not on the moderate emission reductions that can be achieved using existing technologies, but on the breakthrough technologies that are needed to reduce emissions dramatically (see Hoffert et al., 2002, for a technical discussion of some of the technologies discussed here). Of course, the technologies that are ultimately adopted may, and probably will, be very different from the ones discussed here. However, the challenges discussed in this paper are likely to confront any technology that is to play a signiﬁcant role in this effort. As I shall explain, these challenges are formidable; and they are not only or even mainly technical. Indeed, it is possible that the revolution needed to reduce dramatically emissions of greenhouse gases will fail. If it does fail, however, other responses can be expected. The incentive to ﬁnd ways to adapt to climate change will be very powerful, and adaptation will also require technological changes. Should the climate change abruptly, the incentive to “engineer” the climate will be strong. There will be a climate–technology revolution, but its nature will depend on the institutions we develop to address the challenge we face. Scott Barrett 55

Table 1 Benchmark Carbon Prices

($/metric ton CO2)

Benchmark Carbon price ($/tCO2)

Actual values Current global marginal cost 0 RRGI auction price 3 EU Emission Trading Scheme price 30 Estimated values Social cost of carbon Stern (2007) 85 Nordhaus (2008) 7 Optimal carbon tax 2100 (Nordhaus 2008) 55 Cost-effective tax to meet 2°C target In 2010 (Nordhaus 2008) 16 In 2100 (Nordhaus 2008) 235

Note: The estimates from Nordhaus (2008, table 5-1), are in 2005 dollars. The estimate from Stern (2007, p. 344) is in 2000 dollars. “RGGI” stands for Regional Greenhouse Gas Initiative.

Carbon Price

The carbon price, always expressed in this paper in U.S. dollars per metric ton 1 of CO2 , is an important value in the discussion that follows. It will help to have some benchmarks in mind. Table 1 presents eight reference values. The ﬁrst is the current cost of reducing global emissions by one ton, which is approximately equal to zero. The second is the market price of allowances auc- tioned off in late 2008 by the new Regional Greenhouse Gas Initiative (RGGI in the table), which is an agreement among 10 northeastern states to cap and then reduce

CO2 emissions from the power sector. (Six states participated in the ﬁrst auctions: Connecticut, Maine, Maryland, Massachusetts, Rhode Island, and Vermont). The third entry is the price of allowances for carbon emissions traded under the European Union’s Emission Trading Scheme (ETS), also as of late 2008. The remaining values in the table are estimates. Two estimates are presented for the social cost of carbon—the global damage (calculated as a present value) caused by increasing emissions by one ton today, which also represents today’s optimal carbon price. These values, estimated by Stern (2007) and Nordhaus (2008), are strikingly different, due mainly to different approaches to discounting. Nordhaus’s optimal carbon tax increases over time (as will Stern’s), though it remains lower than Stern’s current value through the end of this century. However, Nordhaus’s optimal program allows temperature change to exceed 2°C (as a point estimate; recall that the equilibrium temperature change associated with any con-

1 In the literature, units are sometimes expressed in tons of carbon. One ton of carbon is equivalent to

3.67 tons of carbon dioxide. Prices in $/tC are thus equivalent to 3.67 times the price in $/tCO2. 56 Journal of Economic Perspectives

centration level is uncertain). Nordhaus also calculates the cost-minimizing program for limiting temperature increase to 2°C (again, as a point estimate). The initial value in this sequence ($16/tCO2 in 2010) is still small relative to Stern’s social cost of carbon, but by 2100 the value is much larger than Stern’s initial value

($235/tCO2 versus $85/tCO2). Hence, whether the carbon price should become very large this century depends not only on discounting but also on the perceived need to limit temperature increase to 2°C.

CO2-Free Energy

The most obvious way to reduce emissions (apart from conservation) is by substituting CO2-free energy for fossil fuels. The main alternatives are wind, solar, and nuclear energy. (I discuss bioenergy options later. Geothermal and ocean energy can also play a role, but these options are not discussed in this paper.) Table 2 summarizes the challenges to scaling up use of these technologies. My discussion focuses on the economics of doing this. Wind and solar power have recently attracted investment by venture capitalists, inspiring visions of a “new, new economy.” Google, for example, hopes to develop renewable energy at lower cost than coal (they call their initiative, “RE Ͻ C”). If the new energy entrepreneurs succeed, the market will solve a big part of the challenge without the need for government intervention let alone international cooperation. But is their goal realistic?

Wind Energy Wind energy is already economic on a small scale even at a zero carbon price, and the technology has been improving. Scaling up, however, will require more than the expected improvements in wind turbine technology. DeCarolis and Keith (2006, p. 402) ﬁnd that wind power situated near Chi- cago, the Windy City, can compete with the alternative of natural gas at a carbon price of $38/tCO2. To increase the scale of wind power, the carbon price must be even higher. At a price of $76/tCO2 , wind power can be situated farther from Chicago. Distance from the center of demand increases transmission costs but is valuable because it reduces intermittency (wind speeds at different locations are less correlated) and thus the need for backup gas turbine capacity. Transmission costs currently depend on the installed grid system, which was conﬁgured to suit the existing model of electricity generation—large power plants located close to metropolitan areas. To connect areas where wind power is plentiful to population centers where the electricity is needed requires a new transmission infrastructure relying on high-voltage direct current rather than the existing standard of alter- nating current. The economics of scaling up wind power thus depend on the costs of this complementary technology. Storage is another complementary technology for irregular power sources like wind. If, when the winds blow strongly, surplus wind power is used to produce The Coming Global Climate–Technology Revolution 57

Table 2

Non-CO2 Energy

Wind Solar Nuclear

Viability Wind energy is a proven Photovoltaics are Generation III technology technology. High- proven. Large solar already available. altitude wind power concentrated power Generation IIIϩ and IV has not yet been projects are being under development. demonstrated at scale. planned. Space solar Fusion still at the basic power has not yet science stage. been demonstrated.

Economics Varies by location. Varies by location. Current nuclear technology Depends also on Concentrated solar competitive with coal transmission costs, can compete with and natural gas (for a storage opportunities, fossil fuels in some “high” price) in the U.S.

and costs of sun-rich locations at at just over $27/tCO2

alternatives. $35/tCO2. (MIT, 2003).

Risks/co- Displacement of fossil Risks associated with Operation safety, long-term beneﬁts fuels by wind will beaming power by waste storage, and lower local pollution, lasers or microwaves. proliferation. but there has been Risk of solar satellite local opposition to being attacked. large-scale wind farms, including off shore.

Diffusion Improvements in Depends on Diffusion may be helped by complementary complementary standardization and technologies like technologies. Space development of new, transmission and solar power could be smaller units. energy storage will used to supply power facilitate spread. anywhere.

Scale Wind energy could Available solar energy Nuclear capacity 3–4 times potentially meet all exceeds the world’s current level by 2050 the world’s power total power needs. would displace 3.0–6.6

needs. GtCO2 per year, depending on whether it replaced natural gas or coal (MIT, 2003, p. 3). This is about 5–10% of business-as-usual emissions (IEA, 2008, p. 41). Under this scenario, nuclear’s share of generation rises from 17% to just 19% by 2050 (MIT, 2003, p. 3).

Governance No signiﬁcant Space solar power may Profound challenges, multilateral issues. have military uses; a especially for long-term treaty on its storage, reprocessing, deployment and use and proliferation. may be needed.

Note: GtCO2 is gigatons of carbon dioxide. 58 Journal of Economic Perspectives

hydrogen that is then stored under pressure in a reservoir, then that stored energy could be burned in a turbine (without releasing greenhouse gases) when the winds fail. DeCarolis and Keith (2006, p. 407) estimate that hydrogen storage becomes competitive at a carbon penalty of about $93/tCO2. Here again, local conditions are important. Denmark meets an amazing 17 percent of its electricity needs from wind power, but that is only possible because of electricity trade with Norway (International Energy Agency, 2008, p. 361). When the winds are favorable, Den- mark exports power to Norway, and Norway conserves its hydropower. When the winds fail, Norway releases the energy stored in its dams to produce more electricity for export to Denmark. This inexpensive form of storage is not universally available. Wind power can also be scaled up by building capacity in new places, such as offshore, where the winds are often stronger and blow more consistently than on land. A more radical idea is to capture the energy in the jet stream, ten or so kilometers above the Earth, where the wind is even stronger and steadier. “Flying windmills” could generate power both for lift and transmission, via a cable, to a ground station. Kites could turn a generator as they gained altitude and then have their angle changed so that they descended using less power than they generated in their ascent. Though these technologies have yet to be demonstrated at scale, they could, by some estimates, be competitive even without a carbon penalty (Roberts et al., 2007). There is plenty of wind; according to Roberts et al. (2007, p. 137), wind energy could supply “roughly 100 times the power used by all human civilization.” But practical obstacles remain to be overcome, such as the need to restrict airspace and to bring ﬂying generators down for maintenance and during storms. Also, the jet stream winds are not available everywhere.

Solar Energy Solar energy tends to be abundant in places where wind energy is scarce (high pressure areas have fewer clouds but also less wind). Photovoltaic systems, which convert solar energy to direct current electricity, are already in use, but they operate at low efﬁciency and are only economic in sun-rich off-grid areas. “Concentrated solar power” is a technology that, as the name implies, raises the density of solar energy using mirrors to produce heat, which can then be used to turn a turbine for electricity generation. An example is the “power tower,” a system of sun-tracking mirrors that beam concentrated solar power to a receiver at the top of a tower, through which ﬂows a working liquid for driving the turbine. An advantage of this technology is that it can be scaled to the size of a central power plant (individual units are being designed to generate up to 250 megawatts of power). It can also store thermal energy to produce electricity at night, addressing the problem of intermittency. Solar energy is abundant; the electricity needs of the entire U.S. economy, for example, can be met with concentrated solar power, taking up an area of just 100 square miles (IEA, 2008, p. 379). However, this technology is only suited to areas with intense, direct solar radiation, such as the U.S. Southwest, North and Southern Africa, Australia, and parts of India, China, and Central Asia. (Of course, the power Scott Barrett 59

produced in these places could be shipped to other locations, but this means overcoming the transmission problems discussed previously.) According to a preliminary study by Wheeler (2008, p. 6), concentrated solar power is competitive with coal in Botswana at a carbon penalty of about $35/tCO2. Sub-Saharan Africa needs power to develop, but if it is to develop without increasing greenhouse gas emissions, this carbon penalty will have to be paid. As indicated in Table 1, this price should be paid, if not now, then in the not-too-distant future. However, the prices in Table 1 are global prices. International negotiations will need to resolve the question not only of whether the price should be paid but who should be responsible for paying it. A more radical idea is “space solar power.” This technology would use huge photovoltaic arrays to capture the sun’s energy in space, convert it to direct electrical current, and then beam the electricity to Earth using microwaves or lasers. To produce this energy, solar satellites would be placed in high altitude, geosynchronous orbit, and spaced far enough apart so that at least one unit faced the sun at all times—a solution to the intermittency problem. Macauley and Shih (2007) calculate that, as compared with alternatives such as combined cycle gas turbines and wind, space solar power could be competitive in meeting incremental electricity demand by 2030 in places like California, the U.S. Midwest, Germany, and India—provided fossil fuel alternatives faced a carbon penalty of about 2 $15–25/tCO2. This estimate makes space solar power look very appealing, but it may be optimistic—among other things, the economics of space solar power depend on enhancements in complementary technologies, such as those that can reduce Earth-to-orbit transportation costs. Privately funded R&D can help to improve the prospects for wind and solar power, but breakthroughs are very likely to need a sizable carbon penalty, coupled with widespread investments in complementary infrastructure and technology and basic R&D.

Nuclear Power An expansion of nuclear energy has the potential to reduce greenhouse gas emissions signiﬁcantly and within decades using proven technology. It also has disadvantages. Addressing these will require innovation and institutional changes. The economics of nuclear power are hugely sensitive to capital costs—and, thus, to differences in construction time, plant scale and design, and utilization rates, all of which can vary from country to country. In the United States, as shown in Table 2, nuclear power can compete with coal and natural gas at a carbon 3 penalty of just above $27/tCO2. The economics of nuclear power have generally

2 Macauley and Shih (2007, p. 116) assume a penalty of 1.5–2.5 cents/kWh. They derive their estimates from Krupnick and Burtraw (1996, p. 438), who argue that a climate damage cost of one dollar per ton of CO2 translates into one tenth of a cent per kWh. This puts the carbon penalty at about $15–25/tCO2. 3 The Massachusetts Institute of Technology (2003) study shows that, at this carbon price, nuclear competes with natural gas provided the price of gas is “high” ($6.72/MCF). When I ﬁrst wrote this paper 60 Journal of Economic Perspectives

been unattractive in developing countries, because high capital costs favor scale (most nuclear plants are at least 1,000 megawatts in size). In places with little grid capacity, and in regions with low population density, smaller plants are needed. The so-called “pebble-bed modular reactors” being developed in South Africa will be as small as 100 megawatts. This emerging technology could help to spread nuclear power if, as planned, the diseconomies of its small scale were offset by economies of mass production. Uranium is an exhaustible resource. Just over 400 plants operate now, and according to current estimates, the quantity of reserves remaining can fuel 1,000 new reactors over the next 50 years (Massachusetts Institute of Technology, 2003). Uranium supply, however, is unlikely to limit nuclear power’s expansion this century. As uranium becomes scarcer, prices will rise, creating incentives for new discoveries. Other ways can also be found to ease the supply constraint, including the reprocessing of spent fuel, construction of fast-breeder reactors, extraction of uranium from seawater, and substitution of thorium fuel. The safety of nuclear power plants has improved. Chernobyl, site of the worst nuclear accident ever, was an early generation nuclear plant that lacked a contain- ment structure. Three Mile Island was a more modern plant; though its reactor core melted, no radiation escaped from this facility. Newer nuclear designs incor- porate passive safety features; in the event of an accident, these reactors will shut down automatically. The so-called Generation IV designs now being developed will be safer still. Nuclear waste remains highly radioactive for thousands of years, and solutions to its long-term disposal have yet to be implemented. The obvious long-term storage solution is geological storage, either in repositories like Yucca Mountain in Nevada, several hundred meters below the Earth’s surface, or in boreholes drilled several kilometers deep. So far, however, neither kind of site has been developed, anywhere. Nuclear waste has instead been stored temporarily, on site. The magni- tude of the storage requirement is breathtaking. As noted by the authors of an MIT (2003, p. 10) study, “To dispose of the spent fuel from a steady state deployment of one thousand [1,000 MW ] reactors of the light water type, new repository capacity equal to the nominal storage capacity of Yucca Mountain would have to be created somewhere in the world every three to four years.” New technologies that “parti- tion” and “transmute” the long-lived waste could reduce the storage burden, but these technologies have yet to be demonstrated. Reprocessing extracted plutonium from the spent fuel, making it available in a form that can be handled easily, increases the risk of nuclear weapon proliferation. The production of fresh fuel also poses a risk to proliferation. Most nuclear power today is produced in rich countries. If nuclear power is to play a major role

in 2008, natural gas prices in the United States were higher than this. When I revised the paper in early 2009, prices were lower. In the years since the MIT study was completed, plant construction costs have increased signiﬁcantly. The Coming Global Climate–Technology Revolution 61

in reducing greenhouse gas emissions, fast-growing poor countries will also need access to this energy source. The challenge is to facilitate access without making it easier for more and more countries to master either end of the fuel cycle. This is the purpose of the International Atomic Energy Agency’s proposal to establish a fuel bank of low-enriched uranium. However, to halt proliferation, it will also be essential to reduce the incentive for states to acquire nuclear weapons, and this will require strengthening the existing nonproliferation regime. Nuclear fusion offers a near ideal alternative to ﬁssion—or, indeed, almost any alternative energy source. At least in theory, its fuel is abundant; it poses no risk of a nuclear accident; it yields no high-level waste; and its fuel and waste pose a relatively small risk of weapons proliferation. Fusion power is also a long way from being demonstrated. A new, big science R&D project (the International Thermo- nuclear Experimental Reactor or ITER), currently being constructed in France, is expected to take nuclear fusion to the next stage. However, the future of fusion is speculative. It is not expected to contribute to reducing greenhouse gas emissions before 2050 (IEA, 2008, p. 306). Substituting carbon-free energy for fossil fuels is an obvious way to reduce emissions, but each of the options considered here faces numerous obstacles. Putting a price on carbon will help, but additional policies will be needed to reduce emissions dramatically. Moreover, as we succeed in substituting carbon-free energy for fossil fuels, the price of fossil fuels will fall (all else being equal, of course), lowering the returns to incremental investments in carbon-free energy. For all these reasons, we also need to consider ways of reducing or offsetting the emissions associated with fossil fuel use.

CO2 Capture and Sequestration

Emissions can be reduced at the power plant by removing CO2 before it reaches the air. CO2 can also be removed directly from the air. Table 3 summarizes the economics of, and the challenges associated with, both of these options.

Power Plant Capture and Storage

Fossil fuels can be burned without contributing to climate change if the CO2 that would ordinarily be emitted is captured, transported, and stored somewhere other than the atmosphere. CO2 pipelines already exist, and CO2 can also be transported at sea by ships similar to the ones that now transport liqueﬁed petroleum gases. There are no serious technical obstacles to CO2 transport. The technical challenges concern capture and storage.

CO2 capture can be done either after combustion, by removing CO2 from the stack gases, or before combustion, by removing it from the fuel. CO2 is already captured for certain industrial purposes, like the production of hydrogen for fertilizer manufacture (after being captured, this CO2 is currently released into the air). Other capture techniques are only now being demonstrated. Capture involves Table 3

CO2 Capture and Sequestration Technologies

Carbon Land-based capture/storage biomass Ocean Increasing ocean Industrial air fossil fuel plants capture/storage fertilization alkalinity capture

Viability Many component No such plant Veriﬁcation Not yet Prototypes being technologies are yet difﬁcult; demonstrated. developed. proven, but plant- demonstrated. would likely scale capture/ have to rely storage has not on models. been demonstrated.

Economics $40–90/tCO2 or $50–$110/ $4/tCO2 “No practical $135/tCO2 less (IEA, 2008, tCO2 (see (Kite-Powell; method now exists (Keith, Ha- p. 270); $25–$30 Table 4). see note); for adding Duong, and

(IPCC, 2005, $5tCO2 alkalinity to the Stolaroff, 2005); p. 41); $54–$68/ (Whaley; see ocean at reasonable $100–$200/tCO2 ... tCO2 (Anderson note). cost ” (Stephens (Zeman and and Newell, 2004, and Keith, 2008, Keith, 2008); less

p. 109). Retroﬁts p. 238). than $100/tCO2 more expensive. (Lackner and Sachs, 2005).

Risks/ Geologic storage Storage risks; Main concern Increased alkalinity Sequestration co-beneﬁts poses local environmental is effect on will reduce ocean risks. environmental damage from ocean acidiﬁcation, but risks; ocean shift to ecosystems. there will also be storage may pose biomass. harmful wide-scale risks. environmental Leakage of stored consequences.

CO2 another risk.

Diffusion Diffusion will Diffusion will Can be done Can be done as a Can be done as require either require either as a project. project a project. standards or a standards or a high carbon price. high carbon price.

Scale Geological storage Storage Can sequester Because relies on Can stabilize 1,690–11,100 limited as less than ocean–atmosphere concentrations

GtCO2 (IPCC, above. Land “several mixing, rate of CO2 at virtually any 2005, p. 31); another hundred uptake very slow. level. Ocean storage constraint. million tons 2,000–12,000 of carbon per

GtCO2 (IPCC, year” 2005, p. 35). (Buesseler et al., 2008, p. 162). Governance London No major Parties to two Who decides Who decides the Convention parties multilateral treaties have whether this concentration endorsed carbon challenges. cautioned should be done, level? storage beneath against and on what sea ﬂoor. large-scale scale? fertilization.

Note: The estimates of costs for ocean fertilization are from personal communications with Hauke

Kite-Powell of the Woods Hole Oceanographic Institution and Dan Whaley of Climos. GtCO2 is gigatons of carbon dioxide. Scott Barrett 63

substantial economies of scale, and is appropriate only for large emitting facilities such as power plants. Existing technologies for power systems can capture about

85–95 percent of the CO2 that is produced, but energy is needed to separate the

CO2 , and so the net amount of CO2 captured is closer to 80–90 percent (Inter- governmental Panel on Climate Change, 2005, p. 22).

The economics of capturing CO2 from power plants depend on the type of plant used for comparison purposes (Anderson and Newell, 2004). Costs also depend on whether the plant is designed for capture and storage; retrofits are more costly than purpose-built facilities, not least because existing plants are built close to where electricity is demanded rather than where their CO2 can be stored. China has recently been bringing on line about one new coal-fired power plant every week. None of this capacity has been added with regard to the possible need to retrofit these plants for carbon capture and storage.

CO2 can be stored in geological formations such as oil and gas reservoirs, deep saline aquifers, and un-minable coal beds. Small amounts of CO2 are already being sequestered in places like the Weyburn oil ﬁeld in Saskatchewan and in the Utsira aquifer off the coast of Norway. When stored at depths below 800 meters, CO2 is transformed into a liquid or supercritical state. Supercritical CO2 is more buoyant than water and some crude oils. To ensure that this CO2 remains trapped below ground, it must be stored in formations with impermeable cap rock. Over centuries and millennia, CO2 stored at these depths will dissolve in water, and become denser. Over millions of years, further chemical reactions will convert the stored

CO2 into solid carbonate minerals. Sudden releases, however, are possible in the near term and may pose health and environmental risks (by displacing oxygen in the neighborhood of the release). More gradual releases are also possible and may contaminate groundwater and soils—along with allowing CO2 to escape into the atmosphere.

There is plenty of geological storage capacity for CO2. In recent years, emissions of CO2 have been around 28 billion metric tons per year. Only a fraction of this amount can be captured economically from large sources, and so the capacity estimates shown in Table 3 imply that there is enough geologic storage capacity to lock away at least a century’s worth of CO2 from such sources, and probably much more than that.

The capacity of the deep ocean to store CO2 is greater still, but the oceans are not a truly permanent solution. Ocean mixing would, over a period of centuries, return the CO2 to the surface, where it would be released into the atmosphere very gradually (for the same reasons that the ocean’s surface waters currently absorb

CO2 very slowly). Deep ocean storage would also change ocean chemistry at the injection site and, over time, as the ocean waters mixed, throughout the world’s oceans. The ecological consequences of large-scale, deep ocean storage are unknown.

A ﬁnal possibility is to ﬁx CO2 by accelerating the process of mineral carbon- ation using natural silicates. CO2 stored in this way would have no chance of atmospheric release, and there is enough silicate available to store all the CO2 we 64 Journal of Economic Perspectives

are ever likely to emit. However, this method of storage is costly (carbonization alone would cost about $80/tCO2), and mining the silicate rock would be environ- mentally disruptive (IPCC, 2005, p. 247; Lackner and Sachs, 2005, p. 247). Geo- chemical storage may only be economic, for a high enough carbon price, in regions lacking underground storage sites (Stephens and Keith, 2008). As shown in Table 3, estimates of the costs of capturing, transporting, and storing CO2 underground range from $25/tCO2 to $90/tCO2. Comparing these estimates with the carbon prices shown in Table 1, an economic case can be made either for adopting this technology very soon or later this century. No reasonable case can be made for ignoring this option. An imperative must be to learn more about this technology, including more about its costs and environmental effects. This will require investment in a signiﬁcant number of demonstration projects, in addition to related basic research.

Biomass Carbon Capture and Storage

Tree planting, or “biomass carbon capture and storage,” takes CO2 out of the air. Unlike power plant capture and storage, tree planting need not be coupled to the energy supply system. Biomass (including wood, forestry and crop residues, and various grasses) contains carbon and so can be converted into energy products. Since biomass growth removes CO2 from the atmosphere, substitution of this form of energy for fossil fuels could potentially reduce emissions. Scaling up will be a problem, however, if biomass-for-energy crops displace food crops or result in the destruction of natural ecosystems (though these stresses could possibly be alleviated over time by advances in biotechnology, such as the development of a new strain of algae that could be used to produce a so-called “third generation” biofuel). If biomass were used as a fuel for electricity generation with carbon capture and sequestration, biomass energy would result in negative net emissions. As shown in Table 3, this technology currently costs more than carbon capture and storage from fossil fuel power plants. However, as I shall explain later, energy from biomass may be able to compete, as an offset, with high-cost alternatives for reducing emissions in the transportation sector.

Ocean Fertilization

CO2 can also be removed from the air by fertilizing iron-limited regions of the oceans, to stimulate phytoplankton blooms. If the produced algae (which, like any biomass, absorbs CO2 in the process of photosynthesis) sink into the deep ocean, iron fertilization will remove CO2 from the atmosphere. A related approach is to stimulate phytoplankton blooms by creating an upwelling of nutrients from deep water to the surface in parts of the ocean where the surface waters are low in nitrates. The potential for this kind of air capture is limited. It would work in only a few locations; at best it could play only a small role in stabilizing concentrations. There are also practical problems, such as how to verify the amounts of CO2 sequestered. The Coming Global Climate–Technology Revolution 65

Some experiments in ocean fertilization have been conducted, but to learn more will require large-scale experiments, and these may risk harming ocean ecosystems (Buesseler et al., 2008).

Increasing Ocean Alkalinity

Another way to increase ocean uptake of CO2 is to change the ocean’s alkalinity by adding lime or bicarbonate to the ocean. As noted by Stephens and Keith (2008, p. 228), “If the alkalinity of the oceans was increased, the ocean’s capacity to store dissolved inorganic carbon would increase and the associated increase in acidity would be reduced.” In short, increasing ocean alkalinity could reduce atmospheric concentrations of CO2 even as it reduced ocean acidiﬁcation. The feasibility, environmental impacts, and costs of this approach are unknown, but its role in stabilizing concentrations will in any event be limited because the rate of uptake of CO2 by the oceans is slow.

Industrial Air Capture

CO2 can also be sucked out of the air directly, anyplace on Earth, by means of an industrial process that puts air into contact with a chemical “sorbent,” such as an alkaline liquid. As shown in Table 3, the costs of industrial air capture are expected to be substantial—higher than carbon capture from fossil fuel power plants, because CO2 is more highly concentrated in a plant’s stack gases than in the air. As with all of the new technologies discussed in this paper, these cost estimates are educated guesses. Lackner and Sachs (2005) believe that, with R&D and learning by doing, the costs of air capture might fall to $30/tCO2. Industrial air capture has several desirable features (Sarewitz and Nelson 2008). It would be decoupled from our energy systems, and could be located near geologic sites for long-term carbon storage and away from population areas, where land is cheap. It could also be scaled to any level. Conceivably, every other aspect of the global economy could remain unaltered, and this technology be used to sustain virtually any desired reduction in atmospheric levels of carbon. Pure air capture is a true “backstop technology.” Because it acts directly on reducing concentrations, industrial air capture also offers more options for the timing of investment. Pielke (forthcoming) calculates that air capture could achieve the same concentration target as has been advocated by Stern (2007), for about the same total cost as Stern projected for a scenario of emission reductions. This is because, though air capture has a high marginal cost, its use can be delayed until (for the concentration target advocated by Stern) after around 2050. Even if the intention were not to deploy this technology, it may pay for us to develop it as a hedge against future climate change risks, given its unique ability to be scaled to reduce concentrations directly. Finally, unlike emission reductions, industrial air capture could be deployed by a single country, or by a “coalition of the willing.” This possibility creates an important question for governance: who should decide the level of atmospheric carbon to target? 66 Journal of Economic Perspectives

Transportation Fuels

There are three different options for reducing CO2 emissions in road transport: offsetting the emissions associated with using petroleum-based fuels; using new energy carriers to substitute for these fuels; and developing synthetic hydrocarbon fuels as substitutes.

Offsets Automobile transportation can be made “carbon neutral” by offsetting vehicle emissions with air capture. The great advantage of this approach is that it avoids the need to change the transportation infrastructure. The disadvantage is cost. Row one of Table 4 gives the cost of gasoline and row two the cost of gasoline plus the costs of offsetting the CO2 emissions from gasoline consumption by industrial air capture. These costs are expressed in dollars per gigajoule, a standard unit of energy (one U.S. gallon of gasoline contains about 0.13 gigajoules of energy). The additional cost of the offset, expressed in $/metric ton CO2 , is given in the last column of the table. Air capture is a pure add-on cost; it becomes attractive only if the price of carbon exceeds this value. Recalling the carbon price estimates reviewed in Table 1, this cost is currently not worth paying, though it could become so later this century.

Since biomass growth removes CO2 from the air, use of biomass for electricity production coupled with carbon capture and storage could, like industrial air capture, offset the emissions from gasoline combustion. An estimate of this cost is shown in the third row of Table 4. This option is plainly cheaper than industrial air capture, but the estimates for biomass costs ignore the social costs of this option (such as for land-use change).

CO2-Free Energy Carriers

Hydrogen can carry energy without emitting CO2 when burned. However, hydrogen would require a new infrastructure—for production, storage, and distribution as well as for new vehicles and refueling stations. Powering cars with electricity will also require infrastructure changes—new vehicles with new energy storage systems (batteries), supported by an expanded system for recharging.

Moreover, if the aim were to reduce CO2 emissions, either of these energy carriers would have to be produced using renewable energy, nuclear power, or fossil fuels coupled with carbon capture and storage. Estimates of the costs of these energy carriers, produced by a carbon-free process, are shown in the fourth and fifth rows of Table 4. These costs cannot be compared directly to others in the table. The estimates for hydrogen are for delivery to the vehicle. They ignore the costs of developing an on-board hydrogen storage system and energy conversion. They also ignore the efficiency advantage of fuel cells and the local environmental benefits of reductions in conventional pollutants. Similarly, the costs for electric-powered transportation ignore vehicle costs and the value of differences in Scott Barrett 67

Table 4 Costs of Reducing Road Transport Emissions

Fuel cost (per gigajoule Cost of avoiding emissions from

of energy) gasoline (per metric ton CO2)

Conventional gasolinea $13–$24/GJ — a,b c Gasoline offset by air capture $24–$44/GJ $150–300/tCO2 Gasoline offset by biomass electricity with $16.5–$31.5/GJ $ 50–110/tCO2 carbon capture and storaged Hydrogen made using fossil fuels with carbon $26.5–$41/GJ Vehicle costs additional and will capture and storage be very high.e Electricity made using fossil fuels and carbon $15–$23/GJf Vehicle costs additional and will capture and storage be high; additional infrastructure also required.

Synthetic fuel made from CO2 in biomass $18.5–$21/GJ $80/tCO2; could be negative for high oil price. b,g Synthetic fuels made from atmospheric CO2 $25.5–$34/GJ $150–185/tCO2

Source: Estimates in the fuel cost column are from Zeman and Keith (2008). a Assumes a price of oil of $50–100 per barrel. b Assumes industrial air capture costs of $100–$200/tCO2. c These estimates, slightly rounded, include a 40 percent penalty for the capture and storage needed to offset the emissions associated with reﬁning oil to produce gasoline. d Assumes biomass delivered to the power plant costs $40–$80 per dry metric ton. e According to Keith and Farrell (2003, p. 315), “Costs may exceed [$272/tCO2] if hydrogen cars are to match the performance of evolved conventional vehicles.” Other estimates in the literature are higher than this. f Includes $1–3/GJ to use air capture to offset fugitive emissions from electricity production with carbon capture and storage. g Includes capture and storage costs for the CO2 emitted in the production of hydrogen. vehicle characteristics, such as the number of miles that can be driven between recharges and the onboard space taken up by batteries. The need to match a new infrastructure with a new energy carrier creates a formidable challenge. To make a transition to a “hydrogen economy,” the costs of multiple components need to fall and their performance improve. Given substantial network externalities, such a transition also requires coordinating investment across the different components. An example is the well-known chicken-and-egg problem: consumers will hesitate to purchase hydrogen vehicles unless hydrogen fuel is widely available and competitively priced relative to the alternatives, while energy ﬁrms will be reluctant to build a hydrogen fuel infrastructure so long as hydrogen vehicle ownership is low. Plug-in hybrids could serve as a transition technology. They can run on electricity and gasoline, with the electric power being recharged from the existing grid. In contrast to the all-electric car (given current battery technology), plug-in hybrids can also be driven long distances, making use of the existing refuelling infrastructure. Moreover, as more people drive plug-in hybrids, the recharging infrastructure will grow to serve this new market and vehicle manufacturers will 68 Journal of Economic Perspectives

have an incentive to improve battery performance. These developments will in turn improve the economics of moving to the all-electric car. Finally, as more countries adopt a new automobile standard, there will be incentives for more to do so (Barrett, 2005). As noted before, should the hydrogen or electric vehicle become a new global standard, the challenge will be to ensure that production of these fuels is carbon-free.

Synthetic Fuels A different and more radical approach is to develop new kinds of hydrocarbons that can be distributed and burned using the installed base of capital— synthetic hydrocarbon fuels (Zeman and Keith, 2008). Automobile fuels made today from biomass reduce emissions very little (relative to gasoline); some, such as corn-based ethanol, even increase emissions (Far- gione, Hill, Tilman, Polasky, and Hawthorne, 2008). Advances in plant sciences and “bioreﬁnery” manufacturing may change the economics of biofuels (Ragauskas et al., 2006), but here I focus on a different possibility—using biomass to reduce emissions by separating the carbon in biomass and combining it with hydrogen to make a synthetic hydrocarbon fuel, with the energy needed to power this process being produced from a carbon-free energy source. An estimate of the costs of this alternative is shown in the sixth row of Table 4. These values are very low, but as noted before, they exclude the social costs of biomass production (such as the displacement of food crops). On the other hand, since biomass is not used to power the fuel conversion process, less biomass (and, therefore, less land) is needed to produce a given amount of synthetic fuel as compared with ordinary biofuels.

A related approach would remove CO2 from the air using industrial air capture and use this as a feedstock for producing a synthetic hydrocardon fuel. Essentially, the carbon would be recycled: taken out of the air to make fuel, and then put back in the air as that fuel is burned. So long as the energy that powers this system and that produces the hydrogen input is carbon free, this approach would have an advantage over gasoline offset with air capture—CO2 would not need to be stored underground or in the oceans. The cost of this alternative is shown in the last row of Table 4. Yet another approach is to use the new science of synthetic biology to construct microorganisms that use CO2 as a feedstock to make “fourth-generation biofuels.” These would also be compatible with the existing transportation infrastructure. Overall, there does not yet appear to be an obvious “technological ‘winner’” for the transportation sector (Zeman and Keith, 2008, p. 16). Different futures can be written for the automobile.

Systemwide Effects

I have so far considered technologies in isolation, but systemwide effects will be important. For example, if hydrogen is produced from fossil fuels at a substantial The Coming Global Climate–Technology Revolution 69

scale, the economics of carbon capture and storage will improve. This is because hydrogen production yields a stream of pure CO2 , making capture easier (Ander- son and Newell, 2004, p. 115). Similarly, if the costs of producing electricity from renewable energy or nuclear energy were to fall, the economics of producing hydrogen would improve, as would the returns to R&D into the production of hydrogen by electrolysis. Hydrogen is more appealing as a fuel for maritime than for road transport, but should hydrogen be used for shipping, network effects will improve the economics of using hydrogen for road transport (Farrell, Keith, and Corbett, 2003). Finally, the intermittency problems with certain types of renewable energy like wind or solar matter less if these sources are balanced by complementary generation from hydro, the electricity output of which can be varied relatively quickly, rather than from nuclear or coal with carbon capture and storage (De- Carolis and Keith, 2006, p. 397). These interconnections mean that early policy decisions favoring one technology, for whatever reason, may inﬂuence the evolution of the technology revolution.

Research and Development

None of the breakthrough technologies discussed here can be brought to commercialization without substantial investment in R&D. Unfortunately, as shown in Figures 1A and 1B, energy R&D spending by the members of the International Energy Agency declined after the Framework Convention on Climate Change was negotiated in 1992 and picked up only slightly after the Kyoto Protocol was adopted in 1997. The pattern of R&D spending for parties and nonparties to Kyoto has been very similar. Only in the last few years have countries begun to invest in hydrogen, fuel cells, and carbon capture and storage. This level of R&D spending is too small to kick-start a technological revolution. Why so little R&D? The reason is not only that the knowledge gained from government-funded R&D is a global public good, vulnerable to free riding—that is true for all types of basic knowledge. The bigger problem is that the returns to climate-related R&D depend on the prospects of that R&D leading to reductions in greenhouse gas concentrations—another global public good (Barrett, 2006). These prospects depend on governments cooperating to push up the price of carbon, and so far this effort has been largely unsuccessful. To be sure, there are several international initiatives to promote climate–technology R&D, such as the Carbon Sequestration Leadership Forum (with 21 member countries plus the European Commission) and the International Partnership for the Hydrogen Economy (17 countries plus the European Commission). But these initiatives merely coordinate national activities; they do not increase R&D funding. The ITER (International Thermonuclear Experimental Reactor) project, noted previously, is different; it does involve international ﬁnancing; but ITER will yield substantial beneﬁts unre- lated to climate change. If a better post-Kyoto agreement is not adopted, the incentives to develop and 70 Journal of Economic Perspectives

Figure 1 Energy R&D Expenditure

A: R&D Expenditure for Kyoto Parties 11

10 Conventional 9 Energy

8 Carbon capture & storage 7

4 Nuclear 3 Billions of U.S. dollars (PPP) Hydrogen & fuel cell 2 Renewable energy 1

0 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006

diffuse many of the technologies discussed in this paper will remain weak. The consequences of that failure will be greater climate change, which will stimulate innovation in areas that can substitute for mitigation. These areas are adaptation and geoengineering.

Adaptation

Much adaptation to climate change will be done “automatically” by the market. Other forms of adaptation will involve the supply of local public goods, like dikes, an augmented Thames Barrier, and so on. The incentives for countries to adapt, and to innovate for adaptation, will become very powerful as the effects of climate change appear and become magniﬁed. Indeed, the possibility of adapting to climate change reduces somewhat the incentives to decrease carbon emissions. Poor countries are especially vulnerable to climate change. This is partly because of their geography (Mendelsohn, Dinar, and Williams, 2006). It is also because they lack the private and public sector institutions that would help them adapt to climate change. One area requiring innovation is agriculture. The Con- sultative Group on International Agricultural Research has already begun under- taking research that could reduce future vulnerability dramatically. Examples include heat-tolerant crops, “drought-escaping” rice (varieties that can grow over a shorter cycle), and “waterproof” rice (varieties that survive prolonged ﬂooding). Scott Barrett 71

Figure 1—continued

B: R&D Expenditure for non–Kyoto Parties 11

10 Conventional energy 9

5 Renewable energy 4 Carbon capture & storage

Billions of U.S. dollars (PPP) Nuclear 2 Hydrogen & fuel cell 1

0 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006

Source: ͗http://www.iea.org/RDD/TableViewer/tableView.aspx?ReportIdϭ1͘. Note: Kyoto parties are Austria, Belgium, Canada, Czech Republic (excluded for lack of data), Denmark, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Japan, South Korea, Luxembourg, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, and the United Kingdom. Kyoto nonparties are Australia (which became a party in late 2007), Turkey, and the United States. “Conventional energy” R&D includes energy efﬁciency, fossil fuels (other than carbon capture and storage), and electricity R&D. “Nuclear” includes ﬁssion and fusion. “Renewable energy” includes hydropower, solar, wind, ocean, and bioenergy. “PPP” is purchasing power parity.

These and other innovations will help to reduce the damages from climate change, but there remains a potential for climate change to widen existing inequalities between rich and poor countries.

Geoengineering

It is also possible to compensate for the effect of rising concentrations on temperature. One possibility is to increase the reﬂectivity of the Earth’s surface— for example, by painting roofs white or by planting “shiny crops.”4 Another possibility, with a greater potential to change global temperature, is to reduce the amount of solar radiation that strikes the Earth—that is, to create a planetary parasol.

4 Though tree planning removes CO2 from the atmosphere (resulting in cooling), trees grown in snow-covered tundra areas may reduce reﬂectivity (resulting in warming). What matters is the net effect on temperature. 72 Journal of Economic Perspectives

This could be done at low altitude by “seeding” clouds over the oceans with seawater spray, which would whiten clouds and so increase their reflectivity (their “albedo”). The sea spray could be produced by unmanned, ocean-going vessels, propelled by vertical spinning cylinders that act like sails, with the movement of the ships creating the power needed to produce the spray. Satellites could measure the reflectivity of the clouds, allowing the amount of spray to be varied to maintain the desired level of solar deflection. Preliminary calculations suggest that a fleet of

1,500 vessels could offset the warming effect of a doubling in CO2 concentrations. Though the proposal is still at the drawing board stage, the economics of this proposal are astonishing. According to the designers of this concept, the fleet of vessels needed to offset the effects of a doubling in CO2 concentrations would only cost around $4 billion (Salter, Sortino, and Latham, 2008). The geoengineering proposal that has attracted the most attention so far involves projecting sunlight-deflecting sulfate particles into the stratosphere. The operation would essentially mimic a volcanic eruption, which is known to have a cooling effect. Sulfates are already resident in the stratosphere. To offset the warming associated with a doubling in CO2 concentrations, the amount of sulfates would have to increase 15–30 times (Rasch, 2008). (This multiple may seem high, but it is a very small increase relative to the amount of sulfur in the troposphere.) There are a number of ways to deliver sulfates to the stratosphere: by artillery shells, rockets, high-flying jets, or even hoses tethered to balloons. To maximize the time that the sulfate particles stay in the air, and to create a uniform cooling effect, the particles should be released over the equator. If the intention were to cool only the Arctic, particles could be scattered in the higher latitudes of the northern hemi- sphere (Caldeira and Wood, 2008). Again, the economics of this form of geoengineering seem incredible (Barrett, 2008). Upon reviewing the options in depth, the National Academy of Sciences Panel on Policy Implications of Greenhouse Warming (1992, p. 452, 454) calculated that adding stratospheric aerosol dust to the stratosphere would cost just pennies per ton of CO2 mitigated. Crutzen (2006) thinks the costs would be perhaps $25–$50 billion per year to offset the effects of a doubling in CO2 concentrations, but other estimates are lower (Barrett, 2008). It seems clear that, relative to stabilizing carbon dioxide concentrations, geoengineering is so cheap that cost will not be a major consideration (Keith, 2000, p. 263). So, why not use one of these schemes and forsake the attempt to limit atmospheric concentrations? A number of risks need to be considered: geoengineering would not address the related environmental problem of “ocean acidification”; stratospheric aerosols could destroy ozone; the cooling effect of geoengineering may not preserve the existing spatial distribution of climate; and a geoengineering experiment may have other effects, as yet unimagined. On the other hand, since particles will last at most a few years in the stratosphere (and sea spray only a few days), the geoengineering experiment could be turned off relatively quickly should its effects prove harmful overall. The Coming Global Climate–Technology Revolution 73

A crucial feature of geoengineering is that it could be undertaken unilaterally. Indeed, given this technology’s low cost, a single country may have the incentive to deploy it unilaterally. This makes its provision attractive. It may also create a potential for conﬂict. It is important to distinguish “gradual” from “abrupt and catastrophic” climate change. Gradual climate change may produce winners and losers over the next century. Cline (2007), for example, estimates that a 3°C mean global temperature increase by around the year 2080 would lower India’s agricultural capacity by nearly one-third while increasing capacity in China, Russia, and the United States by perhaps 6 to 8 percent. If India were to deploy geoengineering to avert a local catastrophe, these other countries might complain or intervene militarily or launch a countervailing geoengineering effort to warm the Earth. Over longer periods of time, even gradual climate change would be harmful all around—melting of the Greenland Ice Sheet, for example, would increase sea level by about 7 meters over a millennium. Abrupt climate change, such as a collapse of the West Antarctic Ice Sheet, though unlikely, would have more serious consequences. Should our efforts to limit concentrations fail, or should rapid change occur despite emissions being curtailed sharply, geoengineering may seem worth the risk. Either way, it would seem prudent to be prepared for these possible futures, which is why R&D into geoengineering should begin now. Given the sensitivity around this technology, this R&D should be undertaken cooperatively and openly.

Conclusions

Stabilizing atmospheric concentrations will require fundamental and compre- hensive changes in technology. Market incentives are insufﬁcient to bring about this revolution, and governments generally have weak incentives to intervene unilaterally. International cooperation is needed to set a carbon penalty, to increase R&D spending, to make complementary investments, to coordinate in establishing standards, and to govern the use of new technologies. Many different technologies can help to limit atmospheric concentrations, each with its own advantages and disadvantages, some more speculative than others, and some depending more than others on international cooperation suc- ceeding. It is a cliche´, but true, to say that there is no silver bullet solution to the climate problem. Adaptation and geoengineering do not depend on international cooperation in the same way as the other technological options, but they pose other problems. The different abilities of countries to adapt to climate change may widen existing inequalities. The use of geoengineering may stimulate new international tensions and will not remove all environmental risks. Indeed, it will likely create new ones. Climate change is already underway, and our global institutions and world technology base are starting to co-evolve with it. Given the uncertainties over how quickly and in what ways this mixture of climate, institutions, and technology will 74 Journal of Economic Perspectives

change, a wise policy approach would be to assure that investment is spread over a wide portfolio of possible approaches, some that can be used relatively quickly, some that will not be available until later, and some that are kept in reserve as the twenty-ﬁrst century unfolds. y I am grateful to James Hines, David Keith, Snorre Kverndokk, Lee Lane, Ann Norman, David Popp, Jeremy Stein, Andrei Shleifer, Frank Zeman, and especially Timothy Taylor for comments on an earlier draft, and to Duza J. Baba for research assistance.

References

Anderson, Kevin, and Alice Bows. 2008. “Re- Economics. Available at: http://www.cgdev.org/ framing the Climate Change Challenge in Light content/publications/detail/14090. of Post-2000 Emission Trends.” Philosophical Crutzen, Paul J. 2006. “Albedo Enhancement Transactions of the Royal Society A: Mathematical, by Stratospheric Sulfur Injections: A Contribu- Physical, and Engineering Sciences, 366(1882). tion to Resolve a Policy Dilemma?” Climatic Anderson, Soren, and Richard Newell. 2004. Change, 77(3–4): 211–19. “Prospects for Carbon Capture and Storage DeCarolis, Joseph F., and David W. Keith. Technologies.” Annual Review of the Environment 2006. “The Economics of Large-Scale Wind and Resources, vol. 29, pp. 109–142. Power in a Carbon Constrained World.” Energy Arrow, Kenneth J., Linda Cohen, Paul A. Policy, 34(4): 395–410. David, Robert W. Hahn, Charles Kolstad, Lee Fargione, Joseph, Jason Hill, David Tilman, Lane, W. David Montgomery, Richard R. Nel- Stephen Polasky, and Peter Hawthorne. 2008. son, Roger Noll, and Anne E. Smith. 2008. “A “Land Clearing and the Biofuel Carbon Debt.” Statement on the Appropriate Role for Research Science, 319(5867): 1235–38. and Development in Climate Policy.” AEI Center Farrell, Alexander E., David W. Keith, and for Regulatory and Market Studies, Related Pub- James J. Corbett. 2003. “A Strategy for Introduc- lication 08-12. ing Hydrogen into Transportation.” Energy Pol- Barrett, Scott. 2005. Environment and Statecraft: icy, 31(13): 1357–67. The Strategy of Environmental Treaty-Making. Paper- Hoffert, Martin I. et al. 2002. “Advanced back edition. Oxford: Oxford University Press. Technology Paths to Global Climate Stability: Barrett, Scott. 2006. “Climate Treaties and Energy for a Greenhouse Planet.” Science, ‘Breakthrough’ Technologies.” American Eco- 298(5595): 981–7. nomic Review, 96(2): 22–25. International Energy Agency (IEA). 2008. Barrett, Scott. 2008. “The Incredible Econom- Energy Technology Perspectives in Support of the G8 ics of Geoengineering.” Environmental and Re- Plan of Action: Scenarios and Strategies to 2050. source Economics, 39(1): 45–54. Paris: International Energy Agency. Buesseler, Ken Ol et. al. 2008. “Ocean Fertil- Intergovernmental Panel on Climate Change ization—Moving Forward in a Sea of Uncer- (IPCC). 2005. Carbon Dioxide Capture and Storage, tainty.” Science, 319(5860): 162. Technical Summary. Available at: http://www. Caldeira, Ken, and Lowell Wood. 2008. ipcc.ch/ipccreports/srccs.htm. “Global and Arctic Climate Engineering: Nu- Intergovernmental Panel on Climate Change merical Model Studies.” Philosophical Transac- (IPCC). 2007. Climate Change 2007: Synthesis tions of the Royal Society A, 366(1882). Report, Summary for Policymakers. http://www. Cline, William R. 2007. Global Warming and ipcc.ch/pdf/assessment-report/ar4/syr/ar4_syr_ Agriculture: Impact Estimates by Country. Washing- spm.pdf. ton, DC: Peterson Institute for International Keith, David W. 2000. “Geoengineering the Scott Barrett 75

Climate: History and Prospect.” Annual Review of Ragauskas, Arthur J. et al. 2006. “The Path Energy and Environment, vol. 25, pp 245–84. Forward for Biofuels and Biomaterials.” Science, Keith, David W., and Alexander E. Farrell. 311(5760): 484–89. 2003. “Rethinking Hydrogen Cars.” Science, July Rasch, Philip J., Simone Tilmes, Richard P. Turco, 18, vol. 301, pp. 315–6. Alan Robock, Luke Oman, Chih-Chieh Chen, Keith, David W., Minh Ha-Duong, and Georgiy L. Stenchikov, and Rolando R. Garcia. Joshuah K. Stolaroff. 2005. “Climate Strategy 2008. “An Overview of Geoengineering of Climate with CO2 Capture from the Air.” Climatic Change, Using Stratospheric Sulphate Aerosols.” Philosoph- 74(1–3): 17–45. ical Transactions of the Royal Society A: Mathematical, Krupnick, Alan J., and Dallas Burtraw. 1996. “The Physical, and Engineering Sciences, 366(1856). Social Costs of Electricity: Do the Numbers Add Roberts, Bryan W., David H. Shepard, Ken Cal- Up?” Resource and Energy Economics, 18(4): 423–66. deira, M. Elizabeth Cannon, David G. Eccles, Al- Lackner, Klaus S., and Jeffrey D. Sachs. 2005. “A bert J. Grenier, and Jonathan F. Freidin. 2007. Robust Strategy for Sustainable Energy.” Brookings “Harnessing High-Altitude Wind Power.” IEEE Papers on Economic Activity, no. 2, pp 215–84. Transactions on Energy Conversion, 22(1): 136–144. Macauley, Molly K., and Jhih-Shyang Shih. Salter, Stephen, Graham Sortino, and John 2007. “Satellite Solar Power: Renewed Interest in Latham. 2008. “Sea-Going Hardware for the an Age of Climate Change?” Space Policy, 23(2): Cloud Albedo Method of Reversing Global 108–120. Warming.” Philosophical Transactions of the Royal Massachusetts Institute of Technology (MIT). Society A, 366(1882). 2003. The Future of Nuclear Power. http://web.mit. Sarewitz, Daniel, and Richard Nelson. 2008. edu/nuclearpower/. “Three Rules for Technological Fixes.” Nature, Mendelsohn, Robert, Ariel Dinar, and Larry December 18, pp. 871–72. Williams. 2006. “The Distributional Impact of Stern, Nicholas. 2007. The Economics of Climate Climate Change on Rich and Poor Countries.” Change: The Stern Review. Cambridge: Cambridge Environment and Development Economics, 11(2): University Press. 159–178. Stephens, Jennie C., and David W. Keith. Nordhaus, William. 2008. A Question of Balance: 2008. “Assessing Geochemical Carbon Manage- Weighing the Options on Global Warming Policies. ment.” Climatic Change, 90(3): 217–42. New Haven: Yale University Press. Wheeler, David. 2008. “Crossroads at Mmam- Panel on Policy Implications of Greenhouse abula: Will the World Bank Choose the Clean Warming. 1992. Policy Implications of Greenhouse Energy Path?” Center for Global Development Warming: Mitigation, Adaptation, and the Science Working Paper 140. Base. Washington, DC: National Academy Press. Zeman, Frank S., and David W. Keith. 2008. Pielke, Roger A., Jr. Forthcoming. “An Ideal- “Carbon Neutral Hydrocarbons.” Philosophical ized Assessment of the Economics of Air Cap- Transactions of the Royal Society A. Available at ture of Carbon Dioxide in Mitigation Policy.” http://journals.royalsociety.org/content/ Environmental Science and Policy. b336222382663v17/. This article has been cited by:

1. Michael Hoel, Svenn Jensen. 2012. Cutting costs of catching carbon—Intertemporal effects under imperfect climate policy. Resource and Energy Economics 34:4, 680-695. [CrossRef] 2. James Edmonds, Katherine Calvin, Leon Clarke, Page Kyle, Marshall Wise. 2012. Energy and technology lessons since Rio. Energy Economics . [CrossRef] 3. Johannes Urpelainen. 2012. A model of dynamic climate governance: dream big, win small. International Environmental Agreements: Politics, Law and Economics . [CrossRef] 4. Johannes Urpelainen. 2012. Can strategic technology development improve climate cooperation? A game-theoretic analysis. Mitigation and Adaptation Strategies for Global Change . [CrossRef] 5. Benjamin Jones, Michael Keen, Jon Strand. 2012. Fiscal implications of climate change. International Tax and Public Finance . [CrossRef] 6. Andrew Cheon, Johannes Urpelainen. 2012. Oil prices and energy technology innovation: An empirical analysis. Global Environmental Change 22:2, 407-417. [CrossRef] 7. Johannes Urpelainen. 2012. Costly Adjustments, Markets and International Reassurance. British Journal of Political Science 1-26. [CrossRef] 8. S. Barrett. 2012. Climate Treaties and Backstop Technologies. CESifo Economic Studies 58:1, 31-48. [CrossRef] 9. Edward Nissan, Farhang Niroomand. 2012. Technology diffusion indexes across countries. Journal of Economic Studies 39:1, 31-43. [CrossRef] 10. Johannes Urpelainen. 2011. The strategic design of technology funds for climate cooperation: generating joint gains. Environmental Science & Policy . [CrossRef] 11. Johannes Urpelainen. 2011. Export orientation and domestic electricity generation: Effects on energy efficiency innovation in select sectors. Energy Policy . [CrossRef] 12. Rolf Färe, Shawna Grosskopf, Dimitri Margaritis, William L. Weber. 2011. Technological change and timing reductions in greenhouse gas emissions. Journal of Productivity Analysis . [CrossRef] 13. Johannes Urpelainen. 2011. A California Effect for International Environmental Externalities?. International Interactions 37:2, 170-189. [CrossRef] 14. Gregory F. Nemet, Adam R. Brandt. 2011. Willingness to Pay for a Climate Backstop: Liquid Fuel Producers and Direct CO2 Air Capture. The Energy Journal 33:1. . [CrossRef] 15. Johannes Urpelainen. 2010. Enforcing international environmental cooperation: Technological standards can help. The Review of International Organizations 5:4, 475-496. [CrossRef] 16. David Popp. 2010. Innovation and Climate Policy. Annual Review of Resource Economics 2:1, 275-298. [CrossRef] 17. Lea Kosnik. 2010. The potential for small scale hydropower development in the US. Energy Policy 38:10, 5512-5519. [CrossRef] 18. D.L. Greene, P.R. Boudreaux, D.J. Dean, W. Fulkerson, A.L. Gaddis, R.L. Graham, R.L. Graves, J.L. Hopson, P. Hughes, M.V. Lapsa. 2010. The importance of advancing technology to America’s energy goals. Energy Policy 38:8, 3886-3890. [CrossRef] 19. Scott Barrett, Michael Toman. 2010. Contrasting Future Paths for an Evolving Global Climate Regime. Global Policy 1:1, 64-74. [CrossRef] 20. Isabel Galiana, Christopher Green. 2009. Let the global technology race begin. Nature 462:7273, 570-571. [CrossRef]